empty rules can have attributes

[BFalquet]

enable the usage of attributes with empty_rule to allow things like

> x = c("a", "")
> reformat(x, rule(.to_NA = ""))
[1] a    <NA>
Levels: a

thank you for the review

[github-actions]

🧪 $Test coverage: 97.51%$

Code Coverage Summary

Filename               Stmts    Miss  Cover    Missing
-------------------  -------  ------  -------  ---------
R/assertions.R            54       0  100.00%
R/co_relevels.R           17       0  100.00%
R/cut_by_group.R          15       0  100.00%
R/explicit_na.R           26       0  100.00%
R/filter.R                73       0  100.00%
R/join_adsub_adsl.R       66       8  87.88%   126-133
R/pivot.R                 73       0  100.00%
R/propagate.R             22       0  100.00%
R/reformat.R              58       0  100.00%
R/render_safe.R           38       1  97.37%   30
R/rules.R                 59       2  96.61%   72, 130
R/unite.R                 31       0  100.00%
R/utils.R                 27       0  100.00%
R/zzz.R                    3       3  0.00%    4-6
TOTAL                    562      14  97.51%

Diff against main

Filename        Stmts    Miss  Cover
------------  -------  ------  --------
R/reformat.R       -5       0  +100.00%
R/rules.R          +6      +1  -1.50%
TOTAL              +1      +1  -0.17%

Results for commit: 52b5c4b

Minimum allowed coverage is 80%

♻️ This comment has been updated with latest results

[github-actions]

Unit Tests Summary

    1 files    13 suites   2s ⏱️
  89 tests   86 ✔️ 3 💤 0 ❌
302 runs  297 ✔️ 5 💤 0 ❌

Results for commit 7a06639.

[clarkliming]

looks good. only comment is that it seems we don't need this empty rule class now (from design perspective, empty_rule and other rules are doing the same thing actually)

[BFalquet]

looks good. only comment is that it seems we don't need this empty rule class now (from design perspective, empty_rule and other rules are doing the same thing actually)

New issue opened here. I am merging and the removal to the class will be handled later.